The human gut microbiota refers to the microbial population living within the gastrointestinal tract, which consists of over 38 trillion bacterial cells (Sender et al., 2016) and a relatively minor amount of archaea, eukaryotes and viruses (Qin et al., 2010). Diversity estimates indicate more than 1,000 possible gut microbial species across the human population, with at least 160 dominant species per individual (Qin et al., 2010). Adult gut microbiota composition is shaped by various factors (Jernberg et al., 2007; Khachatryan et al., 2008; Mariat et al., 2009; Scott et al., 2013), with diet potentially modulating over 50% of variation (Zhang et al., 2010; Oriach et al., 2016). In Western contexts, a loss of gut microbial diversity has occurred over time, which may in part be explained by the reduced consumption of dietary fibre and consequent depletion of fermentable substrates for the gut microbiota (Deehan and Walter, 2016). A lack of fermentable substrates can lead to the loss of certain bacterial species, loss of diversity, and reduced production of beneficial fermentation end-products such as short-chain fatty acids (SCFAs) (Sonnenburg and Sonnenburg, 2014). Despite the relative stability of the gut microbiota during adulthood (Wu et al., 2011), short-term dietary changes have been shown to swiftly alter its composition (David et al., 2014). Dietary intervention may therefore represent an appropriate strategy to modulate gut microbiota composition.

Preclinical and human studies have yielded extensive evidence of a “microbiota-gut-brain axis,” which is a bidirectional channel of communication between the central nervous system and gut microbiota (see review Cryan et al., 2019). In humans, depression appears to be linked with an altered gut microbiota composition (Naseribafrouei et al., 2014; Jiang et al., 2015; Aizawa et al., 2016; Kelly et al., 2016; Zheng et al., 2016; Valles-Colomer et al., 2019), though there is little consistency regarding specific differences between depressed and healthy individuals. This may be in part due to dietary variation, as only one study (Kelly et al., 2016) has measured nutrient intake. Faecal microbiota transplantation (FMT) research has indicated a potential causal role for the gut microbiota in depression. After FMT from depressed human patients to microbiota-depleted rats, depressive and/or anxiety-like behaviours have been observed in the recipient animals, yet not in animals who received FMT from healthy controls (Kelly et al., 2016; Zheng et al., 2016). Other research suggests that the gut bacterial-induced inflammatory response from exposure to lipopolysaccharide (part of the external membrane of Gram-negative bacteria), may be implicated in the pathogenesis of depression in humans (Cani et al., 2007; Maes et al., 2012, 2013). Various mechanisms underlying the gut-brain axis have been proposed, with a focus on the interplay between neural, endocrine and immune systems (Cryan and Dinan, 2012).

Gut microbiota composition can be modulated via probiotics, which are live bacteria that provide health benefits when ingested in adequate amounts (FAO/WHO, 2002; Hill et al., 2014). Probiotics enhance host health by competitively excluding and inhibiting the growth of pathogens, triggering cytokine synthesis, and producing SCFAs (Kaur et al., 2002). Specific probiotics that have a beneficial effect on mental health, including those from the Lactobacillus and Bifidobacterium genera (Bravo et al., 2011; Messaoudi et al., 2011; Wang et al., 2016), have been termed “psychobiotics” (Dinan et al., 2013). In one study, 4-week supplementation with Lactobacillus helveticus (strain R0052) and Bifidobacterium longum (strain R0175) led to improvements in depression and somatisation ratings, problem solving skills, and self-blame scores, and reductions in urinary free cortisol (Messaoudi et al., 2011). Concordantly, probiotic-induced improvements in psychological symptoms, predominantly in clinical populations, have been reported in several meta-analyses (McKean et al., 2017; Ng et al., 2018; Goh et al., 2019; Liu et al., 2019; Nikolova et al., 2019; Chao et al., 2020; Zagórska et al., 2020). Other meta-analytic evidence indicates no probiotic effects in clinical or non-clinical subgroups (Reis et al., 2018).

Another method of modulating gut microbiota composition is through the ingestion of prebiotics, which are non-viable substrates that are selectively utilised by host microorganisms and confer a health benefit (Gibson et al., 2017). Prebiotics are partially or completely fermented by microbes in the large intestine, stimulating the growth of certain protective bacteria, especially bifidobacteria and lactobacilli (Cummings et al., 2001). Benefits of prebiotic ingestion include reducing serum endotoxin concentrations and enhancing intestinal integrity, immunity, blood glucose and plasma lipid levels, and absorption of minerals (Iannitti and Palmieri, 2010; Al-Sheraji et al., 2013), as well as the production of SCFAs (Miller and Wolin, 1979). The most established prebiotics are fructans, including fructo-oligosaccharides (FOS) and inulin; and galacto-oligosaccharides (GOS) (Gibson et al., 2017). These compounds are forms of soluble dietary fibre, and are found in varying quantities in fruits, vegetables, grains, legumes, and nuts. As such plant foods are naturally fibre-rich, prebiotic effects may influence the key differences in gut microbiota composition linked with a long-term high-fibre diet (Simpson and Campbell, 2015). Dietary interventions involving an increase in fibre intake and overall diet quality have been reported to reduce depressive symptoms in individuals with clinical depression (Jacka et al., 2017) and elevated depressive symptoms (Francis et al., 2019; Parletta et al., 2019). Preliminary studies using prebiotic supplements to improve mood have yielded some promising findings (Schmidt et al., 2015; Kao et al., 2016; Johnstone et al., 2021), albeit inconclusive overall, likely due to heterogeneity and the limited research base (Desmedt et al., 2019; Liu et al., 2019; Hofmeister et al., 2021).

The “Gut Feelings” trial was designed to extend the existing literature on prebiotic and probiotic supplementation, and test whether consuming a diet rich in high-prebiotic plant foods would improve gut microbiota composition (Tuohy et al., 2012) and, in turn, enhance mental health. Specifically, we developed a high-prebiotic diet and administered it in comparison with a multi-strain probiotic, and a synbiotic combination of both. We were additionally interested in whether synbiotics [the complementary combination of probiotics plus prebiotics (Saulnier et al., 2008; Swanson et al., 2020)] would produce an additive or synergistic effect and, if so, prove more useful than stand-alone probiotics or a high-prebiotic diet. To our knowledge, this was the first study to examine the effects of a high-prebiotic diet on human mental health, addressing the literature gap concerning a lack of whole-dietary interventions targeting brain function and behaviour (Berding et al., 2021b). Our primary aim was to establish whether dietary intake of prebiotics and/or probiotic supplementation would improve mood in adults with symptoms of psychological distress and low prebiotic food intake, relative to placebo. Secondary aims were to measure treatment effects on depression, anxiety, stress, wellbeing, health-related quality of life (fatigue and wellbeing factors), bowel health, and sleep, and whether treatment response was moderated by intervention adherence. We hypothesised that the synbiotic treatment would outperform all others in improving mood, while both prebiotic and probiotic treatments would be superior to placebo, and that this pattern would remain consistent across measures.

A total of 119 participants were randomised to the four study arms (28 to prebiotic diet, 30 to probiotics, 32 to synbiotics, and 28 to placebo). One participant was not included in the analysis of the primary outcome as a technical error via the MHS online assessment centre led to their baseline POMS 2-SF data being unrecoverable. Therefore, 118 participants were included in the analysis. The sociodemographic and clinical features of the sample are summarised in Table 1. Groups were similar on key demographic and clinical variables. The CONSORT chart describing recruitment and follow-up is displayed in Figure 1. Retention in the trial was moderate with 93 (79%) completing the study. There were six non-completers in the prebiotic diet group, seven in the probiotic group, nine in the synbiotic group, and three in the placebo group, which was similar between groups [χ² (3) = 2.85, p = 0.42]. The main reasons for discontinuing were loss of interest/no reason given (n = 7), exclusionary medication use (e.g., antibiotics; n = 6), inability to attend sessions (n = 5), and adverse events (n = 3).

The Gut Feelings trial assessed the effect of a high-prebiotic diet, probiotic supplement, and their synbiotic combination, on mental health symptoms in adults. We found that the high-prebiotic diet improved symptoms on the primary outcome, TMD, relative to placebo. There were no significant effects for other trial arms. Improvements in anxiety, perceived stress, and sleep were also noted in response to the high-prebiotic diet. For the probiotic supplement, an improvement in wellbeing was noted, although this result should be considered preliminary. Unexpectedly, no benefits were noted in response to the synbiotic combination of probiotics and high-prebiotic diet.

We report that adoption of a high-prebiotic diet for an 8-week duration may induce improvements in global mood disturbance and anxiety. Although there is no previous literature specifically on prebiotic-rich dietary interventions and mental health for comparison, a recent meta-analysis by Liu et al. (2019) investigated the effects of isolated prebiotic fibre supplements and probiotic supplements on psychological symptoms. No significant benefit for symptoms of depression (Cohen’s d = −0.08, p = 0.51) or anxiety (d = 0.12, p = 0.11) was found in response to the prebiotic supplements, contrary to our current results for the high-prebiotic diet. Aside from the few studies included in the meta-analysis, there are several potential explanations for this disparity. First, it is possible that isolated preparations of prebiotic fibre could be less effective than a diet high in prebiotic-rich plant foods. Our treatment diet involved consumption of minimally processed high-prebiotic foods across several food groups, promoting intake of a broader range of prebiotic fibres and candidate prebiotics (e.g., short-chain FOS, GOS, inulin, β-glucan, resistant starch, and polyphenols) than most prebiotic supplement formulations, which may induce different fermentation effects (Williams et al., 2017). Second, perhaps the observed mental health benefits from a high-prebiotic diet may be partially attributed to greater vegetable, fruit, and total fibre intake (and consequent changes to dietary macro- and micronutrient composition), as opposed to merely the action of prebiotics (Giri and Sharma, 2022). Indeed, in our study, individuals who followed the high-prebiotic diet not only had significantly higher oligosaccharide and resistant starch intake than non-diet groups at trial completion; they also had greater dietary quality scores than non-diet groups, indicating an overall improvement in their dietary pattern. Further, in natural whole foods, prebiotic fibre occurs in concert with other forms of dietary fibre, minerals, vitamins, phytochemicals, and other bioactive compounds. For instance, oats contain fructans and GOS (Biesiekierski et al., 2011), along with resistant starch, β-glucan, arabinoxylans, sterols, and phenolic compounds (Rose, 2014). Ferulic acid alone, the most commonly found phenolic in whole grains, exhibits anti-inflammatory activity, impacts on redox pathways, and may exert neurological benefits (Neacsu et al., 2017). Given these considerations, it is likely that a high-prebiotic diet facilitates not only gut microbiota-dependent outcomes but also protective effects from dietary improvement in general.

A recent meta-analysis by Firth et al. (2019, 2021) examined studies of dietary intervention on mental health in predominantly non-clinical samples. Relative to inactive control, dietary improvement, which usually includes increasing intake of fibre-rich plant foods, modestly reduced depressive symptoms (Hedge’s g = 0.114; p = 0.035), but not anxiety symptoms (g = 0.077, p = 0.302), although there were fewer available data for anxiety. By contrast, in response to our high-prebiotic dietary intervention, we found relatively stronger evidence of benefit to anxiety symptoms than depressive symptoms, as well as reduced TMD. While potentially a chance finding, this divergence may also reflect a distinct biological or psychological effect of a high-prebiotic diet versus interventions targeting dietary quality more broadly; or it might represent a ‘floor effect’ whereby the mild baseline depression symptoms in our sample reduced potential scope for improvement. A third explanation for why a high-prebiotic diet might yield greater mental health symptom improvements than prebiotic supplements is that supplements are easily blinded and placebo controlled, unlike dietary interventions. The greater effect size observed in the current study for the high-prebiotic diet might then be a consequence of placebo effects or biases related to symptom reporting, such as the Hawthorne effect. On the other hand, if such effects were the driving factor in improving endpoints, this should also have been true of the synbiotic intervention. Notably, the high-prebiotic diet used in our study did not alter gastrointestinal symptoms or BMI at trial completion, therefore its possible psychological benefits cannot be attributed to perceived improvements in bowel function, or weight loss, respectively. The latter finding is consistent with another dietary intervention study that targeted individuals with clinical depression, which achieved improvements in depressive symptoms independent of weight changes (Jacka et al., 2017).

Despite the high-prebiotic diet being associated with a reduction in mood symptoms, there was no difference in the slope of treatment response between the prebiotic and placebo groups from weeks 2 to 8, indicating that much of the apparent diet-induced mood benefit occurred during the first two weeks of treatment. The mood benefit was then sustained, but not increased, relative to placebo, at later visits. A similar pattern was noted for the intake of prebiotic fibre in the diet treated groups, which was maintained at a stable level (considerably higher than the non-diet groups) from weeks 2 to 8 (see Supplementary Table 2). As gut microbiota composition and gene expression can be modulated by dietary interventions as brief as 5-days in duration (2014), such a rapid treatment response achieved within the first two weeks of the study may be biologically plausible. As our week-20 follow-up data indicate (Supplementary Figures 1, 2), the mood effect may be reversible, in line with prebiotic intake dropping back down to near-baseline levels, which may be expected due to the gut ecosystem reverting to its previous composition after cessation of the high-prebiotic diet.

During the trial, participants’ substantially higher prebiotic consumption may have stimulated the proliferation of beneficial gut bacteria such as lactobacilli, bifidobacteria, and butyrate-producing species (So et al., 2018), potentially facilitating protective downstream effects on mental health. Mechanistically, such effects may be partially attributed to SCFAs (primarily acetate, butyrate, and propionate), which are gut microbial metabolites produced from prebiotic fermentation. This process may involve both direct fermentation from prebiotics into SCFAs, and indirect fermentation via cross-feeding, in which fermentation end-products of some microbes are utilised by others to produce SCFAs (Holscher, 2017). SCFAs support gut barrier integrity and intestinal immunity, protecting against peripheral inflammation and ultimately, neuroinflammation (Berding et al., 2021a). Butyrate also directly interacts with the central nervous system, regulating the microglial inflammatory response (Huuskonen et al., 2004; Swann et al., 2019). In line with these SCFA anti-inflammatory effects, meta-analytic evidence indicates that oligosaccharide (e.g., FOS or GOS) supplementation reduces C-reactive protein (CRP) concentrations (McLoughlin et al., 2017). SCFAs can also enhance the blood-brain barrier and modulate levels of neurotransmitters and neurotrophic factors (Berding et al., 2021a). Moreover, SCFAs can reduce the cortisol response to psychosocial stress in humans (Dalile et al., 2020), indicating a modulation effect on the HPA axis. Consistent with this, two prebiotic interventions have produced reductions in cortisol levels (Schmidt et al., 2015; Farhangi et al., 2018). Although our study did not measure faecal metabolites or inflammatory markers to trace potential mechanisms, we plan to analyse changes in participants’ faecal microbiota composition and mental health outcomes in a future publication.

In addition to its beneficial effect on the primary outcome, TMD, the high-prebiotic diet also appeared to produce improvements in perceived stress (assessed by PSS) and sleep (LSEQ) at trial completion. The improvement in perceived stress aligns with studies reporting lowered waking cortisol (a stress hormone) after 3-week GOS supplementation (Schmidt et al., 2015), and reduced perceived stress in a within-group, dose-dependent manner after 4-week consumption of a diet high in both prebiotic and fermented foods (Berding et al., 2022). Our results differ from prebiotic interventions that yielded no effects on stress after consumption of FOS (Schmidt et al., 2015), xylo-oligosaccharides (Childs et al., 2014), agave fructans (Ramnani et al., 2015), or inulin-rich vegetables (Hiel et al., 2019). Variation between studies in duration, measures of stress, sample size, and type of prebiotic may partially explain the differing effects. Our finding of an improvement in sleep in response to the prebiotic diet aligns with promising animal (Thompson et al., 2017) and infant (Colombo et al., 2021) research, yet ours appears to be the first human adult prebiotic trial to report a sleep improvement. Human studies that reported no change in sleep used a mixture of FOS and inulin (Smith, 2005; Buigues et al., 2016), or GOS (Johnstone et al., 2021), with varying trial lengths and populations. Interestingly, recent preclinical literature indicates that gut microbial metabolites produced in response to prebiotic fermentation, such as SCFAs or neuroactive steroids, may play a key role in modulating both stress physiology (Thompson et al., 2017, 2020; van de Wouw et al., 2018) and sleep (Thompson et al., 2017). Hence, to clarify effects on stress and sleep, further human research using a high-prebiotic diet is recommended, particularly with faecal metabolome analysis to enable investigation of potential mechanisms.

Although we found that the probiotic treatment produced an improvement in wellbeing (assessed by SF-36 wellbeing subscale) and weak evidence of improved anxiety (BAI), we emphasise that these findings should be interpreted with caution due to their exploratory nature. There was little evidence of benefit for probiotics on the primary outcome, TMD. Recent meta-analyses appear consistent with our probiotic results in similar populations. Liu et al. (2019) found that probiotics produced a reduction in anxiety and depressive symptoms; however, the effect sizes were small (Cohen’s d for depression = −0.24; anxiety = −0.10). Moreover, the authors found that larger probiotic-induced mental health effects were evident in clinical versus community samples, with the latter non-significant on their own. Other meta-analyses have reported an absence of probiotic effects on mental health in non-clinical populations (Ng et al., 2018; Reis et al., 2018; Goh et al., 2019; Chao et al., 2020), with one exception (McKean et al., 2017). Considering such information, our findings that probiotics may improve wellbeing or anxiety require further replication in larger samples, and for anxiety, our results align with existing data drawn from non-clinical populations. It should also be noted that probiotic effects appear to be strain-specific (McFarland et al., 2018), therefore a different multi-strain probiotic formulation may produce different results to ours.

An unexpected finding of our study was that the observed mood benefits of a high-prebiotic diet were not apparent when combined with a probiotic supplement. This means that we should be cautious in interpreting our findings as evidence that the prebiotic diet was the driver of the improved outcomes observed in the prebiotic intervention group. In the synbiotic group, there was weak evidence of an antagonistic interaction between prebiotic diet and probiotic supplements for TMD (p = 0.063). Specifically, the synbiotic treatment led to an 8.8-point deterioration in POMS 2-SF score compared to the anticipated score had the probiotic and prebiotic treatments been additive, i.e., no interaction present. It is important to note that the antagonistic interaction does not imply that the synbiotic combination of prebiotic diet and probiotics was harmful. Rather, the synbiotic group reached a similar result to the placebo group for TMD at trial completion, which fell far short of the anticipated additive or synergistic effect in a factorial design. Similar weak evidence of antagonistic interactions between prebiotic diet and probiotics were found for secondary outcomes anxiety (assessed via BAI), stress (PSS), sleep (LSEQ), and wellbeing (SF-36 wellbeing subscale). Conclusions regarding the antagonistic interaction, however, should be tempered by the fact that this study was not appropriately powered to detect such an interaction (Simon and Freedman, 1997) and the 95% confidence interval for the interaction effect for the primary outcome was very wide (−0.41 to 18.1).

It is unknown why the synbiotic combination did not appear to result in beneficial effects in line with the prebiotic diet-only intervention, as synbiotics have been reported to stimulate the gut microbiota (Saulnier et al., 2008) and SCFA production (Childs et al., 2014) more than probiotics or prebiotics alone. Further, synbiotics have been shown to improve psychological outcomes (Haghighat et al., 2021; Moludi et al., 2022), reduce inflammatory markers CRP and tumour necrosis factor-α (McLoughlin et al., 2017), decrease oxidative stress (Salehi-Abargouei et al., 2017), and increase levels of serum brain-derived neurotrophic factor (Haghighat et al., 2021). Though random chance cannot be ruled out as an explanation for the differing apparent efficacy of the prebiotic and synbiotic treatments, other explanations may be proposed. First, it may be possible that the ingested probiotic bacteria competed with endogenous microbiota for prebiotic dietary substrates (O’Toole and Cooney, 2008), potentially resulting in less substrate availability for endogenous bacteria such as butyrate-producing species, which may have led to outcomes such as lowered butyrate production. Second, a possible consequence of enhanced stimulation of the gut microbiota via synbiotics may be an increase in bowel symptoms; in such a scenario, treatment benefits may be somewhat diminished by bowel-related discomfort such as bloating. This possibility was investigated in our analysis of the Rome-IV bowel health data. While there was no evidence of worsened bloating, constipation or stool quality, there was suggestive, although inconclusive, evidence that the synbiotic produced an increase in abdominal pain severity relative to placebo (OR = 3.10; p = 0.053). The potential interaction of a high-prebiotic diet and probiotic formula should be evaluated in a larger, future trial, and it is recommended that such studies employ a measure of bowel symptoms to elucidate whether symptoms may mediate any mental health changes.

In summary, this was the first RCT to examine the mental health outcomes of a high-prebiotic dietary intervention, probiotic supplements, and their combination. We found that a diet high in prebiotic-rich whole plant foods may reduce self-reported mood disturbance, anxiety, and stress, and improve sleep in non-clinical adults. A high-prebiotic diet might therefore be a useful strategy for promoting mental health in non-clinical populations. We found limited evidence to support the use of the specific probiotic strain combination used in this study, and no evidence for a synbiotic combination of this probiotic with a high-prebiotic diet, for improving mental health in the current sample. Larger confirmatory RCTs in both clinical and non-clinical populations are needed.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: https://osf.io/ye8b7/.

The studies involving human participants were reviewed and approved by the Melbourne Clinic Research Ethics Committee (TMCREC 289). The patients/participants provided their written informed consent to participate in this study.

JS, TF, FJ, JH, RO, and JR contributed to conception and design of the study. TF, GO, and JS administered the project. TF and RO developed the intervention. TF and GO delivered the intervention. JS and CN acted as supervisors. TF, LC, and N-JM managed data curation and wrote the original manuscript draft. LC and TF performed the statistical analysis. All authors contributed to manuscript revision and approved the submitted version.